Troubleshoot power supply for Dataframe 40.

[olePigeon]

I have an old Supermac Dataframe 40.  It has an old Seagate MFM drive in it (the same time in an Apple HD 20, except 40MBs .... ooooooo)

I'd like to get it running again, but the power supply for the case is borked.

I'm not sure where to start looking.  I was going to replace the capacitors and transformer, but my local part shop didn't have any.  Digikey and Mouser are ridiculously expensive compared to my local shop (or my local shop is ridiculously low-cost.)

Attached is a picture.

Caps look OK.

The symptom is that no power is getting to the two power connectors at the end that connect to the controller card and the hard drive.

 

[olePigeon]

Well, I just ordered a set of capacitors.  $1 per capacitor was the cheapest I could find on Digikey.  Unfortunately my local shop didn't have the right kind, and they're about 3 to 10 cents a capacitor.  In addition, they're having a 50% off sale.

 

[techknight]

A good way to snuff out bad caps in a power supply is the heatgun/hair drier trick. 

If that dont get it going, then there is a cap thats really really really bad, or the problem is elsewhere. 

I would have checked that first, and then if that didnt work, check the opto-isolator and the main line fuse. 

Also its a switched-mode power supply, you cant just use any old cap off the shelf in a local parts store, at least not reliably. the caps need to be of low ESR. 

 

[olePigeon]

Well, if the capacitors I ordered don't work, then that's about the extent of my repair ability.  I don't want to spend $100 to buy a meter, so I can't test opto-somethingoranothers.

Any idea where the fuse is located?  I can check that, anyway.

 

[olePigeon]

Alternatively I could just swap the power supply.  Can anyone take a guess whether or not J8 and J9 are the same voltages as a regular SCSI drive?  Could I just take a power supply from a SCSI case and put a Y splitter on the molex?  My only concern is that the older drive is a different voltage, especially the RLL to SCSI converter board.

 

[olePigeon]

OK, some Googling and I know the HDD is the same power connector.  I haven't plugged in the controller board, though.  I don't want to blow it out if it uses a different pinout.

HDD powers up.

 

[olePigeon]

Oh, nevermind.  Looked on the back of the power supply and both connectors are on the same traces.  So they must be the same voltages.  OK, then.  I'll just use a different power supply with a Y connector.  Should be OK.

 

[olePigeon]

Got it to spin up.  System 7 detects the device, but won't mount it.  Some AltaVisting has shown that I need Dataframe Manager version 4.2 to make it System 7 compatible.  Only version I can find is version 4.1.

 

[olePigeon]

Replacing the caps on the original didn't fix it, unfortunately.  Must be something else.

 

[CC_333]

The component at CR9 looks kinda weird, like maybe there's some rust on one of the legs, but I can't tell for sure because the picture's a bit blurry.

I think it might be some kind of high-wattage diode?

c

 

[olePigeon]

I think it's a diode of some sort.  And, yea, it's a bit rusty.  I tested it with my multimeter and it seemed OK.  Voltage one direction, nothing the other.

 

[techknight]

Replacing the caps on the original didn't fix it, unfortunately.  Must be something else.

Yea, I kinda figured that. at least usually with caps its somewhat alive...

What you should check for is monitor the voltage with a DMM when you first plug it in. if you get a quick spike and collapse to 0, its the opto-isolater bad. 

Otherwise, you have a resistor open, or a transistor open or shorted. It is definitely a component failure at this point.